Friday, 22 June 2012

How to measure ROI for email campaigns

ROI isn’t just about measuring. In times like these, to be brutally honest, it’s about winning. 

ROI (Return on [Marketing] Investment) really needs to be measured in financial terms to make the sort of sense your Financial Director wants to hear, although if you look at ‘value’ as the output, value can be defined in a number of ways. You could for example define value by how likely a customer is to speak highly of you, though it’s almost impossible to establish a financial value to this. Some of the wilder social media agencies will tell you a Facebook “Like” equals £10, which is every bit as useful as a Tweet being the same as a gumboot!

For simple - monetary - ROI, there are two approaches. You can either track your customer all the way to a completed online sale, or track changes in behaviour against a benchmarked control.

You can add a time dimension to this too by looking at the long-term value of a customer - the Customer Lifetime Value (CLTV) - which may be very different to their short-term value. If you’ve ever subscribed to TM Lewin or Richer Sounds emails, where there’s a constant barrage of sales until you capitulate, you’ll know all about balancing an increase in short-term spend against burn-out. In other words, milking the customer too hard usually increases churn... I may buy three times in quick succession but then you’ll lose me forever.

A customer must, by law, have opted in to your email communications, so you know who they are and, provided your email service provider has half-decent tracking and you can tie their purchases to emailed call to action, then you can see the ROI immediately - you spent X on the campaign, which drove Y sales. Presuming you spent less than the extra profit you generated, it’s probably worth doing again. Over time you can then see curves of drop-off in response and optimise your communications accordingly.

Tracking against a control (or baseline) requires a scientific approach and segmentation (perhaps low, medium and high spenders, defined by behaviour, demographics and motivation). This kind of approach will give you the means to track changes in purchase behaviour and attitude towards your brand over time – powerful stuff, In practical terms, if you have a benchmark (for example Tesco data) then you can compare your audience with Tesco’s and factor out TV campaigns and see the actual effect your email campaign is having.

You’ll probably be able to ascribe certain characteristics to a set of customer segments, for example average purchase frequency, average transaction value and CLTV. (For what it’s worth, you’ll nail 80% of everything you need to achieve with segmentation in around six or seven key segments.) If there’s a clear difference between the reactions of two segments to your campaign, for instance if one segment responds by spending twice as much every month, then you can probably guess which segment to spend the most money on.

This approach also tells you where to spend your advertising budget, because obviously you want more people like them... and conversely you can stop spending money on acquiring more people into the lowest performing segments.

Taken only at its face value, ROI is about working out what you got in return for what you spent. But it’s so much more: it leads you from email marketing to eCRM, and that’s a transformational shift. ROI is about gaining an understanding so you can spend less on low profit customers and more on high profit customers. When you do that, you start winning.